DAG Try
AI-generatedSummary
Mark input items as currently executing a DAG branch's try block, tagging each with branch and operation identifiers and setting their status to 'try_running'. On execution failure, emit the input items to a separate error output with a 'try_failed' status and error details.
Inputs
- Branch Label (required) — Unique identifier for the DAG branch; used to label each item to indicate which branch's try block is running.
- Operation Label (required) — Human-readable name for this operation that appears in execution logs.
- Timeout (ms) — Optional timeout in milliseconds; 0 disables timeout. If set, execution is wrapped in a timer.
- Max Retry Attempts — Number of retry attempts on failure; 1 disables retry.
- Retry Delay (ms) — Delay in milliseconds between retry attempts.
- Retry Backoff — Strategy for retry delay backoff: none, linear, or exponential.
- Retry on 4xx Errors — Whether to retry if a 4xx error occurs.
- Retry on 5xx Errors — Whether to retry if a 5xx error occurs.
Output shape
Two outputs: primary 'Success' output with all input items tagged as in-progress 'try_running' state; secondary 'Error' output with input items tagged 'try_failed' on execution error.
If execution succeeds, all items go to 'Success' output with added metadata fields (_dagBranchLabel, _dagOperationLabel, _dagStatus='try_running', _dagTimestamp). If execution throws, all items go to 'Error' output with _dagStatus='try_failed', _dagError message, and _dagTimestamp.